LS Dyna

LS Dyna

Using commands in Explicit Dynamics

    • Magneticfluxquantum



      I'm doing an Explicit Dynamics simulation regarding impact resistance of a ceramic laminate. I'd like to determine the maximum load (e.q. maximum drop height of a standardized sphere) before material failure.

      I'm using the Principal Stress Failure criterion, as this is obtainable experimentally. Extrapolation, however, messes with my result stresses at the crack tips. I'd like to suppress that. Further, it would be great if I could do a multi-step analysis, where the load is adjusted according to the last cycles results, e.g. whether material failure occurred or not.

      I seem to be unable to insert the necessary Command-Objects, as their icon is greyed out. I have tried using snippets, however I'm unsure how they need to be structured. They look like a python script, however I don't know how exactly one calls a function for example.

      Any tips regarding integration of commands are much obliged.

      Thanks in advance


    • Magneticfluxquantum


      A multi-step analysis can be achieved by adding a 'Direct Optimization' object to your workbench. Parameters set in your analysis can be constrained and maximized/minimized. Material failure can be added as an user defined result and set as a parameter.

      The only question remaining is: does stress-interpolation actually affect material failure decisions or is this purely 'cosmetic'?

      Maybe this will help people in the future.

    • Sandeep Medikonda
      Ansys Employee

      The stresses used in the failure criterion should be based on the stresses at the integration points (Gaussian). The GUI might be displaying nodal averaged results by default though.


    • Magneticfluxquantum

      The displayed element stress is not of concern. Material failure occurs when the element quantity (here: stress) exceeds the failure value of said quantity. The subject itself has proven to be quite comprehensive. Maybe I will post an update once I have clarity. I'd consider this post closed for now.

      Anyway, thanks for your time.

Viewing 3 reply threads
  • You must be logged in to reply to this topic.